The models depend on what data you input, and all of them depend on what the people will do. Is a stay at home order in place? How extensive it is and how well is it followed? Depending on the answers the input data changes, and subsequently the output data.
What people don't take into account is that a majority of cases come from just few places, quite small geographically. The current social distancing is trying to prevent a similar spread to other parts of the country.
To give you quick example: if suddenly all US is under strict lockdown, with no movement whatsoever, the virus will likely fade and disappear in a month. Then all models will be wrong again. However, if half the people will ignore the order it, you might as well throw the modeling away.
No it is not. You, and the author, are operating under the hindsight fallacy.
The whole premise is that anyone else could have done better in this emergency that not a single person foresaw. That is a logically unsound argument.
What author? The lady from CDC said it will be bad, Trump and company jumped on her to shut up, that she's saying this because she's a relative of somebody with who Trump had issues with, I think from department of justice (as if that would be hard). A month and half later, we know she was right. And Trump probably knew too but elected to say otherwise, aka lying.
